By the time ConfiguraThor launched in October 2025, my YouTube tutorials had been bringing in leads for a year. This is how that demand got built.
Eyedex started in 2022 as a side project, selling augmented reality to e-commerce stores. No audience, no leads, no brand, no list to email.
Cold outreach was the obvious move and it was also the one that ran out fastest. Every deal needed the same energy as the one before it. So I went the other way and built something that keeps working while I sleep.
In May 2024 I got my first client for a 3D configurator, and that became the focus. In 2025 I turned the work into a product, ConfiguraThor. That page is about building it. This one is about getting it found.
From 2024 I recorded YouTube tutorials on 3D configurators and AR experiences: how they work, the mechanics behind them, and where they fit in the sales funnel of a store. Built around the questions people type when they try to solve the problem themselves.
It took about 5 months before they brought in a lead. From autumn 2024 they brought in about 3 high-ticket leads a month, each worth €1,000 or more.
Google, YouTube and the LLMs. Same buyer, same intent. I restructured the content so answer engines could extract it, added structured data, and tracked citations across ChatGPT, Perplexity and Google AI Overviews.
In one 28-day window, LLMs cited me 2,010+ times. Every one of those is a person asking how to solve their problem and getting my name in the reply, before they have ever seen my site.
Two of the prices I set were wrong, and I did not find that out in a spreadsheet. I found it out in conversations with buyers, in the pause after I said the number.
Changing it twice cost me time. It also gave me the one thing the competition did not have: a price a small brand can approve without a committee, with no subscription attached.
Every finished build became a story with numbers in it, published and pushed on LinkedIn, so the proof was findable long after the project ended.
Content was the main engine. Next to it I built the outbound and lead tooling in-house, with Python, LLMs and Claude Code.
I pointed the same scraping and personalization at a different market: cold email offering SEO services to clinics in Italy. That campaign got a 14% reply rate. Same tooling, a market where the offer was easier to say yes to.
The search side is the same sequence I run for clients under the name OCAVI: find where buyers are searching, look at where competitors already win, work out where you can realistically compete, then build the pages and the videos that get you found and cited. Neuragate is where I learned it. Eyedex is where I ran it on my own money, in the AI era, on a market that did not know I existed.
The tutorials were the part that compounds. They were recorded once, they kept bringing qualified leads every month, and they cost nothing to keep running. That is the difference between buying attention and owning the place where people look for it.
I find out what your buyers type before I write anything. Not what your team calls the product, what a person with the problem searches at 11pm.
I put the product on video where the searching happens. A tutorial that solves a real problem is a sales asset that keeps working more than a year later.
I treat pricing as something you test with buyers, not something you decide alone. I got mine wrong twice and only found out on calls.
I build the growth tooling myself when that is faster: scrapers, enrichment, alerts, small tools. And I say so when a channel is not working, instead of feeding it more volume.
I check whether AI cites you, not only whether Google ranks you. Being the answer someone reads matters more than being the tenth blue link.
If your buyers are searching and finding someone else, that gap is fixable and it is mostly research before it is writing. Three months to install the system, then it keeps working on its own.